VOL. IV.] Coiitribiilions to M^esteni Botany. 41 



carpels three to five, barely surpassing the calyx or even five 

 lines long, much inflated and almost cartilaginous when short, 

 shining and very sparsely hairy, or in the larger ones 

 membranous, flattened, inflated much or little, abruptly acute, 

 not greatly divergent, seeds one or more, broadly obovate, one- 

 half a line long, ovary always densely white-hairy. From the 

 above it will appear that the leaf character, length of carpels 

 and shape of seeds are very variable in the oldest species. 



Neillia opulifolia (I^.) Brewer & Watson, var. mollis. 

 Brewer & Watson; iV. capitata (Pursh), Greene, Pittonia ii, 29. 



My material comes from Oakland and from Duncan's Mills, 

 Cal., and was collected by myself. So far as my speci- 

 mens go the following is true : Leaves broader than in the 

 type, two to two and one-half inches long, and fully as wide or 

 wider, lateral lobes a little larger than in the type, and very 

 rarely is there any evidence of secondary lobes, as is almost 

 always to be found in the other species of Neillia; leaves more 

 pubescent, and more or less cordate at base; carpels vitreous 

 shining, inflated, very sparsely pubescent, shortly acute; seeds 

 lanceolate obovate, incurved or straight; branches not very long 

 nor climbing among the bushes. The corymbs are occasionally 

 proliferous. The only character relied on by Mr. Greene, that of 

 the seeds, proves in my specimens to be valueless, and I cannot 

 see any other good character on which to keep up the species. 

 In some of my specimens at least the seeds are a little narrowed 

 at the apex, but this doubtless is not constant. 



Neillia nwnogyna (Torrey) Greene, Pittonia ii, 30. This is 

 the A^. Torreyi of Watson, etc., in part. It may te advis- 

 able for the present to keep up this species, but there is no 

 necessity for concealing the probable fact that it is only the most 

 reduced form of N. opulifolia. No character that has ever been 

 given it holds except the less inflated pod. Mr. Greene puts this 

 under the head of " carpels indehiscent," but they are dehiscent 

 doubtless when fully developed as that is the case with the variety 

 malvacca {N. malvacea Greene). The form which grows on recks 

 in Colorado is alone sufficiently distinct, but unfortunately the 

 forms growing on better soil and so better nourished differ. The 



